Fair Work Commission bills worker for 'plain wrong' AI advice

Australia's Fair Work Commission ordered sacked ALDI worker Sadnan Khan to pay $1,230 towards his former employer's legal fees after finding he leaned on "plain wrong" AI output as a "quasi-legal advisor" in a failed dismissal challenge, the ABC reports.

Deputy President Michael Easton wrote that "If Mr Khan had properly read his own AI generated replies … he would have known that his case was doomed," and said he had "taken the very rare step of ordering him to pay some of the employer's legal costs, because his unreasonable conduct caused his former employer to incur those unnecessary costs." Khan had been given repeat warnings his case had "no substantial prospects of success."

The tribunal said generative AI was in part responsible for a 40 per cent surge in its cases between 2023-24 and 2024-25. Around 40 per cent of surveyed cases involved litigants using AI, three-quarters of them via ChatGPT, and non-English speakers were twice as likely to use AI as native speakers. Guidance issued Monday will require applicants from October 20 to disclose any AI use, verify facts, authorities and hyperlinks used, and confirm witness evidence.

Monash University law professor Genevieve Grant told the broadcaster that "if those prompts are inexpert, then the results will be as well," pointing to a "sycophantic tendency" familiar to community members. Macquarie University computer science lecturer Gregory Baker, who tells students "if you don't use AI in my class, you will fail," said ChatGPT does "a terrible job" as an oracle and called it "a huge disadvantage in any oral hearing."

The ruling arrives amid a steady run of regulation stories we've been tracking, 259 in the last 90 days, as courts and tribunals push back on unverified generative output in filings. Khan is undeterred: he told the ABC he now plans to run his appeal through "a mixture of Claude and ChatGPT … and two or three [others] to [get a] mixture of different views."